PDF to JPG or PNG: Choosing a Resolution That Is Neither Blurry nor Huge

Text in a PDF is stored as glyphs at coordinates, and vectors are stored as paths. Nothing in the file says how many pixels wide the page is, because a PDF is meant to be sharp at any output size. Turning it into an image is a rendering step that picks a number of pixels per inch of paper, and every dimension follows from that.

For A4 the numbers are worth memorising: at 72 DPI a page is 595 by 842 pixels, at 150 DPI it is 1240 by 1754, at 300 DPI it is 2480 by 3508, and at 600 DPI it is 4961 by 7016. 72 DPI is the historical value one PDF unit is defined against, which is why converters that do not ask you for a resolution quietly produce something that looks soft on any modern screen.

For reading on screen, posting in a chat or dropping a page into a slide deck, 96 to 150 DPI is right: full-page legibility without waste. For printing, or for archiving a page whose fine print matters, 300 DPI is the standard. Go to 400 or 600 DPI only for line art, small technical text, or pages that will be OCRed — and accept that a 600 DPI render has four times the pixels of 300 DPI, so the file is roughly four times larger.

The cost curve is the thing people underestimate. Doubling the resolution quadruples the pixel count, and JPEG or PNG size follows pixel count closely. A 40-page document rendered at 600 DPI is not twice a 300 DPI version, it is four times, and a folder of them can run to gigabytes. If the goal is to look at a document, 150 DPI answers the question; render at 300 only the pages you actually need.

A photograph rendered to JPG is a good trade: small file, no visible difference. A page of text or a diagram rendered to JPG is a bad one, because JPEG's compression works in 8 by 8 blocks and produces ringing around hard edges — the fuzz that appears around letters and thin lines at low quality settings. PNG stores those edges exactly, so text stays clean, at the cost of a larger file.

Pick PNG when the page is mostly text, line art, charts or screenshots; pick JPG when the page is mostly photographs or gradients. Two more details matter. Pages with no background colour are transparent in the PDF, and PNG keeps that transparency while JPG fills it with white — usually fine, occasionally a surprise on dark backgrounds. And if the source is a black-and-white scan, asking for grayscale rather than colour can cut the size dramatically with no loss of usefulness.

Rendering a 30-page PDF produces 30 images, and how they are delivered matters as much as their quality. One image per page is right when you need a specific page for a document or a slide; a single archive of all pages is right when you need the whole thing offline. Bundling them avoids 30 separate downloads, but it also means you cannot preview before committing to the transfer.

Finally, be clear about what you lose. An image of a page cannot be searched, copied or edited — if the goal was to work with the text, the useful conversion is PDF to TXT, DOCX or HTML, which extracts characters rather than pictures. Rasterising is for when the page itself is the thing you want: a preview image, a thumbnail, a page pasted into a design, or a signature block lifted out of a contract.